parser.c

xml开源解析代码.版本为libxml2-2.6.29,可支持GB3212.网络消息发送XML时很有用.

    else
        ctxt->name = NULL;
    ret = ctxt->nameTab[ctxt->nameNr];
    ctxt->nameTab[ctxt->nameNr] = NULL;
    return (ret);
}
#endif /* LIBXML_PUSH_ENABLED */

/**
 * namePush:
 * @ctxt:  an XML parser context
 * @value:  the element name
 *
 * Pushes a new element name on top of the name stack
 *
 * Returns -1 in case of error, the index in the stack otherwise
 */
int
namePush(xmlParserCtxtPtr ctxt, const xmlChar * value)
{
    if (ctxt == NULL) return (-1);

    if (ctxt->nameNr >= ctxt->nameMax) {
        const xmlChar * *tmp;
        ctxt->nameMax *= 2;
        tmp = (const xmlChar * *) xmlRealloc((xmlChar * *)ctxt->nameTab,
                                    ctxt->nameMax *
                                    sizeof(ctxt->nameTab[0]));
        if (tmp == NULL) {
	    ctxt->nameMax /= 2;
	    goto mem_error;
        }
	ctxt->nameTab = tmp;
    }
    ctxt->nameTab[ctxt->nameNr] = value;
    ctxt->name = value;
    return (ctxt->nameNr++);
mem_error:
    xmlErrMemory(ctxt, NULL);
    return (-1);
}
/**
 * namePop:
 * @ctxt: an XML parser context
 *
 * Pops the top element name from the name stack
 *
 * Returns the name just removed
 */
const xmlChar *
namePop(xmlParserCtxtPtr ctxt)
{
    const xmlChar *ret;

    if ((ctxt == NULL) || (ctxt->nameNr <= 0))
        return (NULL);
    ctxt->nameNr--;
    if (ctxt->nameNr > 0)
        ctxt->name = ctxt->nameTab[ctxt->nameNr - 1];
    else
        ctxt->name = NULL;
    ret = ctxt->nameTab[ctxt->nameNr];
    ctxt->nameTab[ctxt->nameNr] = NULL;
    return (ret);
}

static int spacePush(xmlParserCtxtPtr ctxt, int val) {
    if (ctxt->spaceNr >= ctxt->spaceMax) {
	ctxt->spaceMax *= 2;
        ctxt->spaceTab = (int *) xmlRealloc(ctxt->spaceTab,
	             ctxt->spaceMax * sizeof(ctxt->spaceTab[0]));
        if (ctxt->spaceTab == NULL) {
	    xmlErrMemory(ctxt, NULL);
	    return(0);
	}
    }
    ctxt->spaceTab[ctxt->spaceNr] = val;
    ctxt->space = &ctxt->spaceTab[ctxt->spaceNr];
    return(ctxt->spaceNr++);
}

static int spacePop(xmlParserCtxtPtr ctxt) {
    int ret;
    if (ctxt->spaceNr <= 0) return(0);
    ctxt->spaceNr--;
    if (ctxt->spaceNr > 0)
	ctxt->space = &ctxt->spaceTab[ctxt->spaceNr - 1];
    else
        ctxt->space = &ctxt->spaceTab[0];
    ret = ctxt->spaceTab[ctxt->spaceNr];
    ctxt->spaceTab[ctxt->spaceNr] = -1;
    return(ret);
}

/*
 * Macros for accessing the content. Those should be used only by the parser,
 * and not exported.
 *
 * Dirty macros, i.e. one often need to make assumption on the context to
 * use them
 *
 *   CUR_PTR return the current pointer to the xmlChar to be parsed.
 *           To be used with extreme caution since operations consuming
 *           characters may move the input buffer to a different location !
 *   CUR     returns the current xmlChar value, i.e. a 8 bit value if compiled
 *           This should be used internally by the parser
 *           only to compare to ASCII values otherwise it would break when
 *           running with UTF-8 encoding.
 *   RAW     same as CUR but in the input buffer, bypass any token
 *           extraction that may have been done
 *   NXT(n)  returns the n'th next xmlChar. Same as CUR is should be used only
 *           to compare on ASCII based substring.
 *   SKIP(n) Skip n xmlChar, and must also be used only to skip ASCII defined
 *           strings without newlines within the parser.
 *   NEXT1(l) Skip 1 xmlChar, and must also be used only to skip 1 non-newline ASCII 
 *           defined char within the parser.
 * Clean macros, not dependent of an ASCII context, expect UTF-8 encoding
 *
 *   NEXT    Skip to the next character, this does the proper decoding
 *           in UTF-8 mode. It also pop-up unfinished entities on the fly.
 *   NEXTL(l) Skip the current unicode character of l xmlChars long.
 *   CUR_CHAR(l) returns the current unicode character (int), set l
 *           to the number of xmlChars used for the encoding [0-5].
 *   CUR_SCHAR  same but operate on a string instead of the context
 *   COPY_BUF  copy the current unicode char to the target buffer, increment
 *            the index
 *   GROW, SHRINK  handling of input buffers
 */

#define RAW (*ctxt->input->cur)
#define CUR (*ctxt->input->cur)
#define NXT(val) ctxt->input->cur[(val)]
#define CUR_PTR ctxt->input->cur

#define CMP4( s, c1, c2, c3, c4 ) \
  ( ((unsigned char *) s)[ 0 ] == c1 && ((unsigned char *) s)[ 1 ] == c2 && \
    ((unsigned char *) s)[ 2 ] == c3 && ((unsigned char *) s)[ 3 ] == c4 )
#define CMP5( s, c1, c2, c3, c4, c5 ) \
  ( CMP4( s, c1, c2, c3, c4 ) && ((unsigned char *) s)[ 4 ] == c5 )
#define CMP6( s, c1, c2, c3, c4, c5, c6 ) \
  ( CMP5( s, c1, c2, c3, c4, c5 ) && ((unsigned char *) s)[ 5 ] == c6 )
#define CMP7( s, c1, c2, c3, c4, c5, c6, c7 ) \
  ( CMP6( s, c1, c2, c3, c4, c5, c6 ) && ((unsigned char *) s)[ 6 ] == c7 )
#define CMP8( s, c1, c2, c3, c4, c5, c6, c7, c8 ) \
  ( CMP7( s, c1, c2, c3, c4, c5, c6, c7 ) && ((unsigned char *) s)[ 7 ] == c8 )
#define CMP9( s, c1, c2, c3, c4, c5, c6, c7, c8, c9 ) \
  ( CMP8( s, c1, c2, c3, c4, c5, c6, c7, c8 ) && \
    ((unsigned char *) s)[ 8 ] == c9 )
#define CMP10( s, c1, c2, c3, c4, c5, c6, c7, c8, c9, c10 ) \
  ( CMP9( s, c1, c2, c3, c4, c5, c6, c7, c8, c9 ) && \
    ((unsigned char *) s)[ 9 ] == c10 )

#define SKIP(val) do {							\
    ctxt->nbChars += (val),ctxt->input->cur += (val),ctxt->input->col+=(val);			\
    if (*ctxt->input->cur == '%') xmlParserHandlePEReference(ctxt);	\
    if ((*ctxt->input->cur == 0) &&					\
        (xmlParserInputGrow(ctxt->input, INPUT_CHUNK) <= 0))		\
	    xmlPopInput(ctxt);						\
  } while (0)

#define SKIPL(val) do {							\
    int skipl;								\
    for(skipl=0; skipl<val; skipl++) {					\
    	if (*(ctxt->input->cur) == '\n') {				\
	ctxt->input->line++; ctxt->input->col = 1;			\
    	} else ctxt->input->col++;					\
    	ctxt->nbChars++;						\
	ctxt->input->cur++;						\
    }									\
    if (*ctxt->input->cur == '%') xmlParserHandlePEReference(ctxt);	\
    if ((*ctxt->input->cur == 0) &&					\
        (xmlParserInputGrow(ctxt->input, INPUT_CHUNK) <= 0))		\
	    xmlPopInput(ctxt);						\
  } while (0)

#define SHRINK if ((ctxt->progressive == 0) &&				\
		   (ctxt->input->cur - ctxt->input->base > 2 * INPUT_CHUNK) && \
		   (ctxt->input->end - ctxt->input->cur < 2 * INPUT_CHUNK)) \
	xmlSHRINK (ctxt);

static void xmlSHRINK (xmlParserCtxtPtr ctxt) {
    xmlParserInputShrink(ctxt->input);
    if ((*ctxt->input->cur == 0) &&
        (xmlParserInputGrow(ctxt->input, INPUT_CHUNK) <= 0))
	    xmlPopInput(ctxt);
  }

#define GROW if ((ctxt->progressive == 0) &&				\
		 (ctxt->input->end - ctxt->input->cur < INPUT_CHUNK))	\
	xmlGROW (ctxt);

static void xmlGROW (xmlParserCtxtPtr ctxt) {
    xmlParserInputGrow(ctxt->input, INPUT_CHUNK);
    if ((*ctxt->input->cur == 0) &&
        (xmlParserInputGrow(ctxt->input, INPUT_CHUNK) <= 0))
	    xmlPopInput(ctxt);
}

#define SKIP_BLANKS xmlSkipBlankChars(ctxt)

#define NEXT xmlNextChar(ctxt)

#define NEXT1 {								\
	ctxt->input->col++;						\
	ctxt->input->cur++;						\
	ctxt->nbChars++;						\
	if (*ctxt->input->cur == 0)					\
	    xmlParserInputGrow(ctxt->input, INPUT_CHUNK);		\
    }

#define NEXTL(l) do {							\
    if (*(ctxt->input->cur) == '\n') {					\
	ctxt->input->line++; ctxt->input->col = 1;			\
    } else ctxt->input->col++;						\
    ctxt->input->cur += l;				\
    if (*ctxt->input->cur == '%') xmlParserHandlePEReference(ctxt);	\
  } while (0)

#define CUR_CHAR(l) xmlCurrentChar(ctxt, &l)
#define CUR_SCHAR(s, l) xmlStringCurrentChar(ctxt, s, &l)

#define COPY_BUF(l,b,i,v)						\
    if (l == 1) b[i++] = (xmlChar) v;					\
    else i += xmlCopyCharMultiByte(&b[i],v)

/**
 * xmlSkipBlankChars:
 * @ctxt:  the XML parser context
 *
 * skip all blanks character found at that point in the input streams.
 * It pops up finished entities in the process if allowable at that point.
 *
 * Returns the number of space chars skipped
 */

int
xmlSkipBlankChars(xmlParserCtxtPtr ctxt) {
    int res = 0;

    /*
     * It's Okay to use CUR/NEXT here since all the blanks are on
     * the ASCII range.
     */
    if ((ctxt->inputNr == 1) && (ctxt->instate != XML_PARSER_DTD)) {
	const xmlChar *cur;
	/*
	 * if we are in the document content, go really fast
	 */
	cur = ctxt->input->cur;
	while (IS_BLANK_CH(*cur)) {
	    if (*cur == '\n') {
		ctxt->input->line++; ctxt->input->col = 1;
	    }
	    cur++;
	    res++;
	    if (*cur == 0) {
		ctxt->input->cur = cur;
		xmlParserInputGrow(ctxt->input, INPUT_CHUNK);
		cur = ctxt->input->cur;
	    }
	}
	ctxt->input->cur = cur;
    } else {
	int cur;
	do {
	    cur = CUR;
	    while (IS_BLANK_CH(cur)) { /* CHECKED tstblanks.xml */
		NEXT;
		cur = CUR;
		res++;
	    }
	    while ((cur == 0) && (ctxt->inputNr > 1) &&
		   (ctxt->instate != XML_PARSER_COMMENT)) {
		xmlPopInput(ctxt);
		cur = CUR;
	    }
	    /*
	     * Need to handle support of entities branching here
	     */
	    if (*ctxt->input->cur == '%') xmlParserHandlePEReference(ctxt);
	} while (IS_BLANK(cur)); /* CHECKED tstblanks.xml */
    }
    return(res);
}

/************************************************************************
 *									*
 *		Commodity functions to handle entities			*
 *									*
 ************************************************************************/

/**
 * xmlPopInput:
 * @ctxt:  an XML parser context
 *
 * xmlPopInput: the current input pointed by ctxt->input came to an end
 *          pop it and return the next char.
 *
 * Returns the current xmlChar in the parser context
 */
xmlChar
xmlPopInput(xmlParserCtxtPtr ctxt) {
    if ((ctxt == NULL) || (ctxt->inputNr <= 1)) return(0);
    if (xmlParserDebugEntities)
	xmlGenericError(xmlGenericErrorContext,
		"Popping input %d\n", ctxt->inputNr);
    xmlFreeInputStream(inputPop(ctxt));
    if ((*ctxt->input->cur == 0) &&
        (xmlParserInputGrow(ctxt->input, INPUT_CHUNK) <= 0))
	    return(xmlPopInput(ctxt));
    return(CUR);
}

/**
 * xmlPushInput:
 * @ctxt:  an XML parser context
 * @input:  an XML parser input fragment (entity, XML fragment ...).
 *
 * xmlPushInput: switch to a new input stream which is stacked on top
 *               of the previous one(s).
 */
void
xmlPushInput(xmlParserCtxtPtr ctxt, xmlParserInputPtr input) {
    if (input == NULL) return;

    if (xmlParserDebugEntities) {
	if ((ctxt->input != NULL) && (ctxt->input->filename))
	    xmlGenericError(xmlGenericErrorContext,
		    "%s(%d): ", ctxt->input->filename,
		    ctxt->input->line);
	xmlGenericError(xmlGenericErrorContext,
		"Pushing input %d : %.30s\n", ctxt->inputNr+1, input->cur);
    }
    inputPush(ctxt, input);
    GROW;
}

/**
 * xmlParseCharRef:
 * @ctxt:  an XML parser context
 *
 * parse Reference declarations
 *
 * [66] CharRef ::= '&#' [0-9]+ ';' |
 *                  '&#x' [0-9a-fA-F]+ ';'
 *
 * [ WFC: Legal Character ]
 * Characters referred to using character references must match the
 * production for Char. 
 *
 * Returns the value parsed (as an int), 0 in case of error
 */
int
xmlParseCharRef(xmlParserCtxtPtr ctxt) {
    unsigned int val = 0;
    int count = 0;
    unsigned int outofrange = 0;

    /*
     * Using RAW/CUR/NEXT is okay since we are working on ASCII range here
     */
    if ((RAW == '&') && (NXT(1) == '#') &&
        (NXT(2) == 'x')) {
	SKIP(3);
	GROW;
	while (RAW != ';') { /* loop blocked by count */
	    if (count++ > 20) {
		count = 0;
		GROW;
	    }
	    if ((RAW >= '0') && (RAW <= '9')) 
	        val = val * 16 + (CUR - '0');
	    else if ((RAW >= 'a') && (RAW <= 'f') && (count < 20))
	        val = val * 16 + (CUR - 'a') + 10;
	    else if ((RAW >= 'A') && (RAW <= 'F') && (count < 20))
	        val = val * 16 + (CUR - 'A') + 10;
	    else {
		xmlFatalErr(ctxt, XML_ERR_INVALID_HEX_CHARREF, NULL);
		val = 0;
		break;
	    }
	    if (val > 0x10FFFF)
	        outofrange = val;

	    NEXT;
	    count++;
	}
	if (RAW == ';') {
	    /* on purpose to avoid reentrancy problems with NEXT and SKIP */
	    ctxt->input->col++;
	    ctxt->nbChars ++;
	    ctxt->input->cur++;
	}
    } else if  ((RAW == '&') && (NXT(1) == '#')) {
	SKIP(2);
	GROW;
	while (RAW != ';') { /* loop blocked by count */
	    if (count++ > 20) {
		count = 0;
		GROW;
	    }
	    if ((RAW >= '0') && (RAW <= '9')) 
	        val = val * 10 + (CUR - '0');
	    else {
		xmlFatalErr(ctxt, XML_ERR_INVALID_DEC_CHARREF, NULL);
		val = 0;
		break;
	    }
	    if (val > 0x10FFFF)
	        outofrange = val;

	    NEXT;
	    count++;
	}
	if (RAW == ';') {
	    /* on purpose to avoid reentrancy problems with NEXT and SKIP */
	    ctxt->input->col++;
	    ctxt->nbChars ++;
	    ctxt->input->cur++;
	}
    } else {
        xmlFatalErr(ctxt, XML_ERR_INVALID_CHARREF, NULL);
    }

    /*
     * [ WFC: Legal Character ]
     * Characters referred to using character references must match the
     * production for Char. 
     */
    if ((IS_CHAR(val) && (outofrange == 0))) {
        return(val);
    } else {
        xmlFatalErrMsgInt(ctxt, XML_ERR_INVALID_CHAR,
                          "xmlParseCharRef: invalid xmlChar value %d\n",
	                  val);
    }
    return(0);
}

/**
 * xmlParseStringCharRef:
 * @ctxt:  an XML parser context
 * @str:  a pointer to an index in the string
 *
 * parse Reference declarations, variant parsing from a string rather
 * than an an input flow.
 *
 * [66] CharRef ::= '&#' [0-9]+ ';' |
 *                  '&#x' [0-9a-fA-F]+ ';'
 *
 * [ WFC: Legal Character ]
 * Characters referred to using character references must match the
 * production for Char. 
 *